ELECTRONIC STRUCTURE AND MOLECULAR ASSOCIATION OF

Publication Date: September 1960. ACS Legacy Archive. Cite this:J. Am. Chem. Soc. 82, 18, 5005-5006. Note: In lieu of an abstract, this is the article...
0 downloads 0 Views 265KB Size
COMMUNICATIONS TO THE EDITOR

Sept. 20, 1960

H I

WD)

I C&-C-CHI I

DW)

I + *C&-C-CHa

H(D)

traces of styrene present initially and from progressive oxidation during reaction. This mechanism clearly explains why disproportionation of ethylbenzene is so much more rapid than t ~ l u e n e . ~ The recent isolation of 1-(ethylphenyl)-1-phenylethanes from the treatment of ethylbenzene with ferric chloridee also may be pertinent. The mechanism previously has been postulated to account for some disproportionation results in the presence of fairly high concentrations of hydrogen acceptors.' The present results show that the mechanism can be operative under normal transalkylation or disproportionation conditions and may be general for primary alkylbenzenes; there seems little question that transalkylation of secondary and tertiary alkyl groups generally involves essentially free alkyl cations. A number of obvious corollaries of the new mechanism have not yet been studied. The present results compel that the new mechanism be given serious consideration in other disproportionation and transalkylation studies.

5005

X-ray analysis of the sublimed solid6 indicates intermolecular associations. The material is known to form a dihydrate readily, suggesting a strong tendency for the Ni(I1) to become octahedrally coordinated. m'hile the visible spectrum and magnetic properties of the complex in hydrocarbon solvents exclude a tetrahedral arrangement, octahedral coordination by means of polymerization is not excluded. Molecular weight determinations in dichloromethane indicate that I is substantially polymerized.6 The preparation and characterization of the complexes reported offers additional evidence for this view. Bis - (2,2,6,6- tetramethyl - 3,5- 1ieptanediono)-Ni (II), 11, (R1 = Rz= -C(CH3)3, R3 = H) was found \

c=o

t o be diamagnetic. Hydrocarbon solutions of the material exhibit a single absorption peak in the visible a t 535 mp, with a molar extinction coefficient of 60, giving a red coloration. The material readily forms a blue-green paramagnetic dihydrate. Anal. Calcd. for NiCzzHa04: C, 62.14; H, 8.92. Found: C, 61.98; H , 8.92. Bis - (2,2 - dimethyl - 3 , s - heptanediono) - Ni(II), I11 (R1= Rz = -CH(CH3)?, R3 = H ) , was found t o be paramagnetic, peff (297.4' K.) = 3.41 B.M., in the solid phase. Toluene solutions of the anhydrous material indicate that the magnetic moments of the solutions are concentration and (6) P. Kovacic, C. Wu and R. W. Stewart, THISJOURNAL, 82, temperature dependent. The color of the solu1917 (1960). tions changes from green near 0' to red a t around (7) H. Pines and J. T. Arrigo, ibid., 80, 4369 (1958); L. Schmerling, J. P. Luvisi and K.W. Welch, ibid., 81, 2718 (1959). 50°, corresponding to the growth of a band a t 535 DEPARTMENT OF CHEMISTRY ANDREWSTREITWIESER, JR. m p . All attempts to explain the magnetic properUNIVERSITY OF CALIFORSIA ties of these solutions and the growth of the 535 BERKELEY 4, CALIFORNIA LIANEREIF mp band by means of a Boltzmann distribution of RECEIVED JULY 8,1900 the molecules in the singlet and triplet states2 failed. The concentration dependence of both the moment and the intensity of the 535 mp band ELECTRONIC STRUCTURE AND MOLECULAR excluded this assumption and led to an attempt to ASSOCIATION OF SOME BIS-(8-DIKETONE)NICKEL(1I) COMPLEXES correlate the data assuming a temperatureSir: dependent distribution among monomers and Recently's2 i t was shown from ligand field argu- polymers in solution. Assuming the monomer ments that one may expect planar, tetracoordinate t o be diamagnetic and to be the only species complexes of Ni(I1) to be either paramagnetic or absorbing a t 535 mp, i t was possible to fit diamagnetic depending on the strength of the sur- the data a t several concentrations and temperarounding field. Bis-acetylacetonato-Ni(I1) has tures to a set of equilibrium expressions. been cited as a prominent example of a planar paraAnal. Calcd. for NiCleH3004: C, 58.57; H, 8.11. magnetic material. Studies in this laboratory make Found: C, 58.48; H , 8.29. i t increasingly apparent that the spin-free ground The spin-paired ground state in I1 can be atstate may be caused by intermolecular associations, tributed to the fact that the bulky t-butyl groups instead of being characteristic of the free mono- prevent intermolecular association. Scale molecumeric molecule. lar models indicate that only slight strain is inWhile bis-(acety1acetonato)-Ni(II), I, is planar3 (5) G. J. Bullen, Nature, 177, 537 (1956). Drs. R. Mason and P. and monomeric4 in the vapor phase, an incomplete Pauling at University College London recently have re-examined (1) G. Maki. J . Chem. Phys., 28, 651 (1958); 29, 162 (1959);

as,1129 (1~59).

(2) A. H. Liehr and C. Ballhausen, THISJOURNAL, 81,538 (1959). (3) S. Shibata, Bull. Chem. SOC. J a p a n , 30,753 (1957). (4) F.Gach. Monafsh.Chemic, 21,QS(1900).

Bullen's three-dimensional data and find that the molecules in the trimer are so arranged that each nickel ion is surrounded by six oxygen atoms in a slightly distorted octahedron (private communication). ( 6 ) F. A. Cotton and R. H. Soderberg, t o be published.

5006

COMLKNICATIONS TO THE EDITOR

curred when the Ni(I1) atom in I11 is placed near the oxygens or chelate ring of a neighboring molecule. The tendency for Ni(I1) to become octahedrally coordinate apparently overcomes this strain in the solid. Preliminary studies on a third compound, bis(3-phenyl-2,4-pentanediono)-Ni(II),I V , (R1 = R2 = CH3, RB = CsH5) gives additional support to the idea that the paramagnetism in anhydrous B-diketone complexes of Ni(I1) can be attributed to intermolecular interactions. Models indicate that the phenyl group in IV cannot rotate into coplanarity with the chelate ring due to the proximity of the methyl groups. Close intermolecular association of the Ni(I1) atom with a neighboring molecule would be hindered by the phenyl groups perpendicular to the chelate ring. The anhydrous material, from drying the blue-green hydrate, is red and diamagnetic, a single band being observed in the visible spectrum of the solid a t 535 mp. I n toluene the material appears to behave similarly to 111, the color of the solution changing from green to red as the temperature is increased from 0 to 50”. It also appears that a green crystalline modification of the material can be formed by melting the red powder. Studies are also being carried out on the visible spectrum of I in hydrocarbon solutions a t fairly high temperatures. I t has been observed visually that the color of solutions of I in dibenzyl become reddish-brown near 200’ and reversibly return to the green color upon cooling. Financial support was provided by the U. S. Atomic Energy Commission under Contract No. AT(30-1)-1965 and by the Research Corporation to whom grateful acknowledgment is made. J.P.F. wishes t o acknowledge a Fellowship received from the Allied Chemical and Dye Corporation, SemetSolvay Division.

TTol. 82

the elaboration of the tetracyclines. Therefore we have used the term cosynthesis for this phenomenon to distinguish i t from the previously reported microbial systems involving cross-feeding of essential nutritional factors (syntrophism).2,3 TABLE I COSYKTHESIS OF TETRACVCLIXE ANTIBIOTICS Antiassay bacterial c of

Principal MutantsQ grown alone

tetracycline product of parent strain b

Tetracyclic product of mutant b

7-C1-TC PIUS 7-Cl-5a( 1 l a ) dehydro T C IV- 5 i-C1-TC Unknown E-504 7-Cl-6-demetligl-TC Unknown T-219 TC Unknotvii S.2242 7-Cl-TC Unknowii S-1308

7-Cl-TC

fermented mash, g.11.

0 20

Si1 Nil Si1 Si1 Antiassay b ac teri a1 of

Mutants grown a s mixtures

+ + + + + IV-5 + S-2242

Tetracycline products cosynthesized b

mixed fermentations, d

&!./I.

S-1308 K-5 ifr-5 E-504 ‘r-219 s-2242 5-1308 E-504 E-504 T-219

7-Cl-TC 2.i 7-C1-6-demethyl-TC 0.22 7-C1-TC plus T C 0.74 None Nil 7-Cl-TC plus 7-C1-60,34 demeth yl-TC Xone Til a Some of these mutants were obtained through the TC = courtesy of Dr. J. Growich of these laboratories. tetracycline. S.aureus turbidimetric assay calculated 3 s 7-clilorotetracycliiie. d Corrected for assay of mutants grown alone.

The mechanisms involved in cosynthesis are assumed to be analogous to those involved in syntrophism, that is, the transfer of biosynthetic intermediates from the mutant of the later block to the mutant of the earlier block or, alternatively, DEPARTMENT OF CHEMISTRY J. P FACKLER, JR. the transfer of the missing facility from either MASSACHUSETTS INSTITUTE OF TECHXOLOGY CAMBRIDGE 39, MASSACHUSETTS F. A. COTTON mutant to the other, resulting in the coijperative biosynthesis of the normal product of the paient RECEIVED J c r . ~12, IO00 strain. 14-e wish to report in greater detail the results COSYNTHESIS OF TETRACYCLINES BY PAIRS of our study of one of these cosynthetic pairs. OF STREPTOMYCES A UREOFACIE,VS The two mutants used were coded W-5 and SMUTANTS. 1308; both were isolated directly from high 7 Sir : chlorotetracycline-producing parent strains. In the course of our studies on the biosynthesis Xutant W-5 was an unpigmented mutant lacking of the tetracyclines we have sought for and found the ability to elaborate significant quantities of numerous blocked mutant strains of Streptomyces 7-chlorotetracycline and presumably blocked early aureofaciens Duggar which have lost their ability in the biosynthetic pathway. The other mutant of to elaborate any substantial quantity of tetra- this pair, S-130S, was pigmented; it produced 7 to the extent cycline antibiotics. In order to determine whether chloro-5a(1la)-dehydr~tetracycline~ such mutants might show biological cooperation of about 4.0 g./l. together with about 0.2 g./’l. in the synthesis of tetracycline antibiotics, they of 7-chlorotetracycline. When 48 hour-old shaker were then grown as pairs in mixed fermentation, flask cultures of W-5 and S-1308 were mixed in and the resulting fermented mashes examined for equal volume and incubated an additional 72 tetracyclines. Typical data summarized in Table hours, 2.8 g,/l. of 7-chlorotetracycline and corI show that in many cases substantial quantities respondingly less of ’i-chloro-5a(1la)-deliydroof one or more tetracyclines were produced in the tetracycline were produced. mixed fermentations even when each mutant (1) J. J . Goodman, M. Lfatrishin, R. W. I’ouni( a n d J R r). grown alone produced no appreciable amount. McCormick, J . B a c t . , 7 8 , 4 9 2(1959). (2) J. Lederberg, i b i d . , 62, 503 (1946). It should be noted that the methods used resulted ( 3 ) B D. Davis, Experienlia, 6,41 (1980). in the selection of only those mutants which grew (4) J. R. D. McCormick, P. A . Miller, J . A. Growich, A?. 0 . Sjowell in a corn steep medium’ ordinarily used for lander and A. P. Doerschuk, Tms J O U R N A L , 80, 5572 (19.58)